What do these investors know that we don’t? Are they just too deep in now to quit?
I get that AI is extremely useful.
But the idea that OpenAI can 1) create a much more powerful superhuman AI and 2) also keep the way of making it secret from competitors long enough to make all the money seems doubly unlikely.
What they are seeing is insane user and revenue growth. It's that simple. ChatGPT has like a billion users, and it came out 3 years ago. And then AI is getting better every year in a way that internet companies like Google or Facebook didn't. Google peaked very early in terms of usefulness, then it just spent decades monetizing it. AI is not even funded by ads, people are actually paying for it. It's really hard to get people to pay for something. I don't know if the investors will get a nice return, but given the numbers they're seeing I understand why they're doing it.
Part of me thinks that nobody wants to be the one that pops the bubble, so it's safer to just play along and throw in a few hundred billion here and there.
From watching talks by the various US think tanks (CSIS, CFR, Hudson, etc), the common theme whenever the topic of AI is brought up is that the US is in strategic competition with China where it essentially boils down to economic superiority where the idea is to have US technology spread globally and not China’s. Oh and military application of AI.
I have a hunch the hyper won’t slow down anytime soon if these groups who have the eyes and ears of the government are suggesting to go full steam ahead in the technology.
AI is currently summoning market activity during what should be a recession. Now this bubble isn't only filled with hot air, it's also filled with sewage.
When anybody wonders why everything is getting worse and more expensive, remember that all the money to create new better products is being redirected to this clusterfuck.
What could we have if all that massive amount of money was used in one other field. (Or many others, there is too much money to invest effectively in just one thing)
Title edited by me from "OpenAI won’t make money by 2030 and still needs to come up with another $207 billion to power its growth plans, HSBC estimates"
What do these investors know that we don’t? Are they just too deep in now to quit?
I get that AI is extremely useful.
But the idea that OpenAI can 1) create a much more powerful superhuman AI and 2) also keep the way of making it secret from competitors long enough to make all the money seems doubly unlikely.
What they are seeing is insane user and revenue growth. It's that simple. ChatGPT has like a billion users, and it came out 3 years ago. And then AI is getting better every year in a way that internet companies like Google or Facebook didn't. Google peaked very early in terms of usefulness, then it just spent decades monetizing it. AI is not even funded by ads, people are actually paying for it. It's really hard to get people to pay for something. I don't know if the investors will get a nice return, but given the numbers they're seeing I understand why they're doing it.
Part of me thinks that nobody wants to be the one that pops the bubble, so it's safer to just play along and throw in a few hundred billion here and there.
My theory is they will keep funding it until IPO then cash-out.
From watching talks by the various US think tanks (CSIS, CFR, Hudson, etc), the common theme whenever the topic of AI is brought up is that the US is in strategic competition with China where it essentially boils down to economic superiority where the idea is to have US technology spread globally and not China’s. Oh and military application of AI.
I have a hunch the hyper won’t slow down anytime soon if these groups who have the eyes and ears of the government are suggesting to go full steam ahead in the technology.
FOMO must play a massive role in it.
what else is there to invest in?
AI is currently summoning market activity during what should be a recession. Now this bubble isn't only filled with hot air, it's also filled with sewage.
Short-term gains are great, though!
When anybody wonders why everything is getting worse and more expensive, remember that all the money to create new better products is being redirected to this clusterfuck.
What could we have if all that massive amount of money was used in one other field. (Or many others, there is too much money to invest effectively in just one thing)
The EU also has rampant inflation even though we have no AI sector to speak of.
It seems to me they have realised they can continue printing money to keep stealing from us forever and nobody seems to care.
Inflation in the EU is at 2.5% (2.1% in the Eurozone), that's hardly "rampant".
Ah yes, the official numbers. If only food and rent were going up only by 2.5%
Can they afford a cash-burn marathon against Google/Meta/Microsoft/Ali though?
Not to mention that Google has infinite video data, scraping data (via Search), their own Quantum computers, their own TPUs, their own Cloud....
Gemini seems to suck majorly, I don't know why (it lies a lot and doesn't follow instructions) but I've been wondering about this too for a while
Ali?
Alibaba, maybe
Title edited by me from "OpenAI won’t make money by 2030 and still needs to come up with another $207 billion to power its growth plans, HSBC estimates"
My rule of thumb in these estimates:
"Always double the amount presented in these estimates"
This includes large construction projects, especially government run projects.
So I say 500 billion USD will be needed :)
Thanks grilledchickenw and xfour, fixed
I think you may want to adjust your units there. 500M is a Tuesday breakfast for these AI companies burn rate
500 billon, you mean.
[dupe] Large discussion: https://news.ycombinator.com/item?id=46058065